A triangular convergent mixed methods design was applied to obtain a more thorough understanding of influencing factors for prescribing confidence of final-year medical students in the emergency setting of STEMI. An online survey using the voluntary sampling method and in-depth interviews using purposive sampling were performed simultaneously, analysed separately and then integrated into the final analyses (Figure 1).^9-11 Both quantitative and qualitative data were collected during the same period of study and targeted to achieve the objective of the study and facilitate triangulation and credibility. Qualitative methods were required due to the limited data and comprehensive appreciation of their confidence and competence for a prescription for a case with STEMI. To enrich the interpretation of findings, the relationships between the qualitative and quantitative results were described.

The study was conducted between November 1, 2020, and April 30, 2021, in one single center to minimise confounding effects from different environments, clinical rotation experience, and basic medical training programs. Ethical approval was considered by the Institutional Review Board of Srinakharinwirot University, Thailand (SWUEC-332/2563E). Written informed consent was obtained from all final-year medical students who participated in the study. Participation in the study was free, and all data were presented anonymously.

Missing data are unavoidable when data are collected using questionnaires. A web-based self-administered questionnaire was used in this study to diminish respondents’ stress and the item nonresponse as a result of the respondent’s mistakes or refusals. Some items in the questionnaire tested recall and could be answered ‘do-not-know’, for instance, a question about the dosage of fibrinolytic agents. Therefore, those data can be missing completely at random, and a complete case analysis was used to handle it.¹⁶

Categorical variables are shown as frequency and percentages. The levels of confidence for prescribing antiplatelet and fibrinolytic agents were scored from 1 to 5 according to the 5-point Likert scale and classified into two groups: low- and high confidence based on their means.¹³ The Likert scale responses of their confidence in prescription were analysed using parametric statistics¹⁷ and were provided as mean (M) with standard deviation (SD) and compared with paired t-test.

The influencing factors of their confidence in prescribing both medications were expressed as frequency and percentages in case they also agreed. Those were compared between low- and high-confident groups using a chi-squared test (χ²) or Fisher's exact test. The p-values were two-sided, and a value of less than 0.05 was considered a significant difference. Data were analysed using STATA version 16.1.

During survey data collection, in-depth interviews were conducted in the same period. Purposive sampling was used to recruit the study participants. For inclusion, all participants (25-27 years, n=20) who were final-year medical students and graduates in the academic year of 2020 at the time of the interview and agreed to the audio recording of interview sessions were enrolled. All participants were invited to an in-depth interview during their responses to an online survey. Therefore, they were able to participate and verify in both approaches, and remuneration was not given. The interview concentrated on the interviewees’ experience, self-reported confidence in diagnosis and prescription, and their attitude to barriers to prescribing antiplatelet and fibrinolytic agents for patients with STEMI.

Four trained female study investigators (P.I., P.T., M.K., and V.S.) conducted interviews face-to-face (n=15) in private rooms located within the hospital area or by telephone (n=5) using a semi-structured interview list which consisted of open- and closed-ended questions, initially developed in English and then translated into Thai (Appendix). Each interview was individually operated by one study staff who was randomly assigned to five participants. The written informed consents were collected, and interview data were audio-recorded by one mobile phone, which was only used in this study. Study numbers without personal identifiers were uniquely assigned for each subject. All participants who also responded to the online survey included 10 (50%) female and 10 (50%) male students with a mean age of 26.0 (SD=0.8) years. The median time of interviews lasted 18.02 (IQR:16.71- 21.35) minutes, and they were conducted in the Thai language.

Initially, all four interviewers were trained, reviewed and elucidated on the probing technique for the first five interviews. Then they applied to collect insight information from the interviewees. Conversations were transcribed verbatim in Thai and translated to English after checking by T.A. and they were not returned to participants for comment. All transcription of the data was performed manually without the use of a database.

An inductive thematic analysis was performed following steps documented by Braun and Clarke.¹⁸ To understand the factors affecting their confidence in prescribing antiplatelet and fibrinolytic medications, the interview transcripts were reviewed, and initial codes were generated, with searching, reviewing and naming for themes. Saturation of themes was achieved during data collection by all interviewers in discussion together. All discrepancies were resolved by consensus with the author (T.A.).

To evaluate in what ways the results of prescribing confidence factors converged or diverged, the attitude of confidence factors provided and free text responses in the questionnaires were compared with themes derived from the qualitative interviews.

Seventy-one respondents (77.2%) correctly diagnosed a case scenario of STEMI, while 83 (90.2%) and 64 (71.9%) respondents accurately prescribed the type and dosage of antiplatelets, respectively (Table 1). Nearly all (n=86, 93.5%) responded to recognising the fibrinolytic agents in treatment STEMI, although some data collected on type and dosage of prescribing fibrinolytic agents were missing, 12 (13.9%) and 57 (66.3%), respectively. Seventy-four (86.0%) and 25 (80.6%) of them could correctly answer the type and dosage of fibrinolytic agents, respectively. The self-reported confidence score of both prescriptions varied in range from 1 to 5, and median of both of them was 3.0. The mean self-reported confidence score of prescribing antiplatelet and fibrinolytic agents were medium, in which the former was significantly higher than the latter (M=3.3, SD=1.1 vs M=2.8, SD=1.0, t ₍₉₁₎ =5.39, p<0.01). The very confident score (score=5) was rarely reported for both antiplatelets (n=13, 14.1%) and fibrinolytic agents (n=1, 1.1%) prescriptions. Their attitude of confidence factors in prescribing antiplatelet and fibrinolytic agents were ranked as follows: experience with prescription, 62(67.4%) vs 64(69.6%); confidence in correctly diagnosis of STEMI, 61(66.3%) vs 61(66.3%); supervision during prescription, 58(63.0%) vs 61(66.3%); and teaching by attending physicians and/or residents, 46(50.0%) vs 50(54.3%), respectively (data not shown).

Table 2 demonstrates no significant difference between low- and high-confident groups of prescribing antiplatelets and fibrinolytic agents with regard to their gender, medical ward practice, their ability of diagnosis STEMI and prescribing type of those medications. Nonetheless, their competence for accurate prescribing dosage of antiplatelets were notably higher in high-confident group [low- vs. high-: 29 (56.8%) vs. 35 (85.4%), χ²_{(1, N =92)}= 9.85, p=0.007], while there was no significant difference for prescribing type and dosage of fibrinolytic agents [low- vs. high-: 6 (18.2%) vs. 20 (33.9%), p=0.25, N 35, Fisher's Exact Test]. By comparing between low- and high-confident groups of their perspectives of prescribing confidence in both medications, previous experience of prescription was significantly recognised in the former, while the correct diagnosis of STEMI was considerably reported in the latter. Teaching and supervision during prescription by their instructors and/or residents were additionally listed in which teaching in class was considerably reported in high-confident students for prescribing fibrinolytic medications [low- vs high-: 13 (39.4%) vs 37 (62.7%), p=0.03, N 92, Fisher's Exact Test]. However, supervision during the prescription of both medications were not differently reported as for their confidence factors between the two groups.

Table 3 illustrates the characteristics of 20 final-year medical students and their interview data. Twelve of them had trained in medical wards and encountered STEMI at medical wards with a median case of 5 (IQR: 4-15), while 11 of them had experienced 1-2 cases of STEMI presented at the emergency room. When compared with prescribing antiplatelets, they tended to have lower confidence in prescribing fibrinolytic agents.

Thematic analysis demonstrated two domains (influencers and builders of confidence) and four themes: (1) experience, (2) knowledge, (3) providing guidelines and (4) mentoring (Table 4). They were derived from the theme codes of recall, own experience, anxiety, fear, teaching, peer-review, approval, consult, clinical practice, guidelines, standing order, and electrocardiography (ECG). Mentoring was frequently mentioned as their confidence builder for the prescription. The details of the four themes are:

(1)  Experience refers to their own experience with the real cases in aspects of diagnosis of STEMI and prescription.

Own experience with the real cases was recognised as having an impact on their confidence in prescribing antiplatelet and fibrinolytic agents. The following statements are offered verbatim as examples:

“I’m still anxious because it's my first attempt to fully manage these on my own.” (No.6, Female, 25 years)

“I’m studying in a medical ward where I have been practising reading ECGs with the STEMI cases. Therefore, I’m quite sure when I see the cases fit with their ECG.” (No.12, Female, 27 years)

(2)  Knowledge relates to recall, teaching in class, and self-directed learning. This theme is defined by the ability of all the participants to diagnose cases with STEMI correctly and recognise medication regarding the type, dosage, adverse effects, and cautions to use. One participant said:

“I’m anxious about the safety of my patients if I can't remember their dosages accurately.” (No.9, male, 27 years)

To increase self-confidence in accurately prescribing both medications, self-directed learning which was related to reading, review and practice, teaching in class, and searching on the internet, were quoted. They also recognised the importance of accurately reading an ECG for diagnosis of STEMI and case-based teaching. Some mentioned:

“Case-based teaching helps me in managing the patient with STEMI by myself” (No. 8, male, 26 years)

“Practice reading ECGs accurately is essential.” (No.10, male, 25 years)

“Self-review of the lessons would help me remember better.” (No.13, Female, 26 years)

(3)  Providing guidelines was themed from adhering to clinical practice guidelines and standing order. It was reported as both a confidence influencer and builder on their prescriptions. One of them stated:

“Whereas I somewhat know, but I’m not certain, I’d rather prescribe medication as stated in the standing order.” (No. 14, Female, 26 years)

(4)  Mentoring was quoted by approval, consultation, and ECG recognition of cases with STEMI. Some mentioned that to increase self-confidence in accurately prescribing both medications, they needed to be supervised by their instructor and/or a senior doctor during their management and prescription. Some of them described:

“In my opinion, mentoring and approval of my diagnosis of STEMI and prescription order of essential medications by teachers and/or residents could build up my confidence in prescription.” (No. 2, male, 26 years)

“In life-threatening cases such as STEMI, I would like my staff come and see with me” (No.8, male, 26 years)

“It would be helpful if someone could observe and assist me while I’m reading the ECG of STEMI.” (No.11, Female, 25 years)

The attitude of confidence factors for prescribing essential medications in STEMI of final-year medical students, which was derived from surveys and interviews, converged. Their experience, knowledge and mentoring were reported in both approaches when only providing guidelines was complementarily found from the qualitative method.

The authors would like to thank Stephen John Pinder (native speaker) and Nattakrit Tongpoonsakdi for conducting a comprehensive English language review on this manuscript, and all last-year medical students of Faculty of Medicine, Srinakharinwirot University and Maneerat, Pinngernor, our academic officer, for their valuable support of this research work.

The authors declare that they have no conflict of interest.